akiiwa-games/AgentCompass / skills/agent-compass
agent-compass
FlaggedChoose owner, model and effort, outbound authorization, advisor use, and continuation for delegated work. Reads local usage and effective account policy, profiles the task with the configured System-One scorer, then allocates a compatible worker. Use for independent worker or advisor work; ordinary local execution needs no routing.
references/level-3-conversion.json
View on GitHub{
"schema": "agent-compass.level-3-conversion/1",
"conversions": {
"2": {
"applies_to": "user-scores.json schema_version 2 (v0.6.7 and v0.6.8), carried to schema 3",
"rule": "A level L at or below 2 is unchanged (levels 1 and 2 are the panel minimum and median, as before). A level above 2 becomes 2 + (L - 2) * (A - 2), A being the level the old level 3 (the anchor claude-opus-5-5) has on the facet scale of the version named in `source`; the segment above 3 continues with the same slope. A category entry of several facets (coding, tool_operation) above 2 is written per facet. Converted levels are written to 2 decimals. A facet whose old scale was a plateau (`above_3_slopes`) gives 2 for every level in (2, 3] and, above 3, 2 + (L - 3) * s: the old scale put level L above 3 at t2 + (L - 3) * (t2 - t1), which is level 2 + (L - 3) * (t2 - t1) / (t3 - t2) on the scale of `scale`.",
"source": {
"kind": "maintainer",
"as_of": "2026-11-18",
"statement": "v0.7.0 references/five-axis-scores.json (schema agent-compass.five-axis-scores/6): `anchor_levels` is the level of z = 0 on each facet scale (t1, t2, t3 = panel minimum, median, maximum, listed under `scale`). They are frozen here so that a local file upgraded on any later version gets the same levels.",
"note": "tool_operation.responsiveness: the scale of schema 2 had t2 = t3 (the anchor sat at or below the panel median), so every level in (2, 3] meant the panel median and a level above 3 went above it by (L - 3) * (t2 - t1); its anchor level is 2.0 by that definition, not read from the scale, and its slope above 3 is (t2 - t1) / (t3 - t2) of the declared scale."
},
"anchor_levels": {
"creativity.creative_writing": 3.0,
"coding.execution": 3.0,
"coding.faithfulness": 2.2358,
"aesthetics.design": 2.9175,
"tool_operation.responsiveness": 2.0,
"tool_operation.image_recognition": 2.9266
},
"above_3_slopes": {
"tool_operation.responsiveness": 1.1018
},
"scale": {
"creativity.creative_writing": [
-3.1448,
-1.9575,
0.0
],
"coding.execution": [
-2.6706,
-1.1906,
0.0
],
"coding.faithfulness": [
-2.3305,
-0.2414,
0.7824
],
"aesthetics.design": [
-2.9457,
-1.3577,
0.1221
],
"tool_operation.responsiveness": [
-1.8994,
0.0106,
1.7441
],
"tool_operation.image_recognition": [
-2.7593,
-1.1851,
0.0939
]
}
}
}
}
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |